A new take on my UBMQ

[jhitesma]

Reprinted the center in MakerGeeks "Bomber Jacket Brown" PETG...which I'd probably just call Black but it has a very subtle reddish/brown tint to it if you print it thin and hold it up to a light.



Did the center in 60% infill and the arm pockets at 15% The center is still a bit more flexy than I'd like...but still the top/bottom plates are were it should get most of it's strength from.

It looks good. The latest tweaks to the design resulted in screw holes that M3 screws fit in perfectly and the arms don't poke out the bottom anymore.



The wires had just barely enough slack to re-assemble it without having to unsolder anything. If anyone builds one of these - leave yourself a bit of slack so you can replace arms/printed parts easily!



I still need new top/bottom plates before I can fly it. Just finished reassembling the MPCNC tonight but still have to mount the motors, wire the steppers and add the belts to it before I could start testing it and then I'll need to print a mount for my dremel before I could try using it to cut new plates. So will probably just cut some quick and dirty temporary ones on the bandsaw tomorrow.

Meeting up with someone to fly tomorrow afternoon so fingers crossed I can have it back in the air by then!

 

[jhitesma]

Top and bottom plates made from some 1/16" aircraft ply...should have used 1/8" but...wasn't sure if my arm screws would be long enough then and this was on top of my pile...worst case...it breaks again. Best case...it works and is lighter



Gave it a quick test flight and still flys well on the same tune even though it's pretty windy today and the weight went down a little to 455g (it was 463 before.)

Going to do some flying with some friends this afternoon, look forward to see what they think of it!

 

[jhitesma]

So I only got to fly one pack and only experienced one "soft" crash into tall grass and one decent smack into the ground ... oh and my first mid-air collision when my friend flying FPV hit me (we both stayed in the air though!) But so far the PETG center is holding up just fine. The top 1/16" ply cracked though so I'm guessing the center flexed and I'm nervous about damaging the PDB/FC if that happens again.

So it seems the PETG is a thumbs up, but the 1/16" ply is a thumbs down. About what I expected.

But the MPCNC is alive and working. So far I've only mounted a pen and used it to do some simple plotting.

First test was the same penguin I had used to test my mini-CNC I built about a year ago:


Since that worked first try I stepped up to something a bit more complex that my daughter requested:


I'm still getting used to working without endstops and it took me a few tries to get the Z movements figured out. First try on the left the paper wasn't totally flat and my starting position was off so the pen didn't contact some spots. Second try in the middle I had misconnected the Z motor so there was no z and the pen dragged. Third try on the right I had the Z too far the other way and smashed the tip of the pen...it worked but mangled the tip and was like fingers down a chalkboard.

The thing is wherever it is when the controller resets is [0,0,0] and I have it set to "machine" 1mm and raise 2mm for moves. So it drops 1mm from where I start to make lines but moves 1mm up from where I start between lines. The other issue is the pen mount isn't ideal. The pen just rests in a cup and is held down with some rubber bands. This gives it some slop to adjust for inaccurate Z and less than 100% level bed...but...it also means the pen and jiggle so I get movement of the pen that lowers accuracy. But I'm making progress. Last night I finally got a good "Elsa" but didn't get a photo of it.

Still I'm feeling confident enough that I printed up this mount last night:


I need to remake the upper hoop though. This was designed for a newer Dremel 3000 not my 15 year old Model 395 Type 5. Thankfully the upper hoop bolts on so I don't have to reprint the whole thing. I also picked up some cutter bits for my Dremel so I should soon be able to start experimenting with machining materials and in a few days will hopefully be able to use it to cut my lexan plates!

 

[C0d3M0nk3y]

Third try on the right I had the Z too far the other way and smashed the tip of the pen...it worked but mangled the tip and was like fingers down a chalkboard.

One might say your CNC needs to "Let it Go".

 

[Twitchity]

Have you thought about getting a 300watt spindle instead of using the Dremel? I upgraded my Shapeoko2 to a spindle and was much happier. The Dremel got worn out fast, but it was also a generic brand so that could have played into it as well. Either way, the 300watt spindles are much quieter than the Dremel, and not too much more expensive. You'll need an external power supply though.

 

[jhitesma]

The dremel was on hand and free I don't plan on doing any serious milling with it. Cutting out plates from 1/8" Lexan is about the most I'll be asking of it. Maybe a bit of 3D engraving on some soft pine. Most people run a 660 deWalt trim router on the MPCNC since they're only about $50 on amazon and there are lots of mount designs for them, plus they're strong enough to mill just about anything the MPCNC can handle. Noise isn't a big concern for me. Yeah, right now it's on my dining room table. But long term I plan on swapping the rails for the 3'x4' set I made up initially and then putting it out in my office (That's why after initially "finishing" it I tore it apart and re-printed everything in PETG - my office only has A/C when I'm out there and I worried that PLA would deform.)

 

[jhitesma]

Well, after talking to Twitchity about cutter bits I placed an order from Drillman1 on ebay and today they arrived.

10 of the bits Twitch uses to cut CF (for some future projects) and 2 "O-flute" bits one 1/8" and one 1/16" that are claimed to be THE bits for cutting "soft plastics" (which includes acrylic and polycarbonate/lexan.)

So over lunch I chucked up the 1/8" O-flute and gave it a go on some acrylic. I started with the 1/8" bit since I was kind of rushing and figured the 1/16" bit would be a lot easier to break if I screwed up.

Well...I did screw up...but thankfully didn't break any bits

For one thing the gcode I ran wasn't optimized for the bit at all, so it tried to do outlines of the screw holes instead of plunge them - doh. Plus I started with my Z a bit too low and as soon as I turned on the dremel it made a nice hole...oops. And finally I rushed "clamping" my workpiece and just taped it down with some blue tape...which let loose halfway through the first perimeter cut...double doh.

Even so the results have me excited:



Amazing the difference a proper bit of tooling makes! It makes a MUCH bigger mess with those fine chips...but the cut quality is astounding! No more chipping and cracking it went through the acrylic beautifully and will probably handle higher feed rates and deeper cuts than what I used as a starting point just fine.



At the top you can see a failed test (forgot to tighten the gantry to the rollers and it shifted while cutting!) with the cheap dremel diamond coated engraving bit I used originally because it was all I could find at a local store. Notice how milky, chipped and rough of a cut it was? The lower piece is the one I cut with the O-flute - the edge is gorgeous and should flame polish up beautifully.

Hopefully tonight I'll have time to redo the gcode and give this a go on some lexan to see if I can make final plates!

 

[C0d3M0nk3y]

10 of the bits Twitch uses to cut CF (for some future projects)

Are you going to cut carbon fiber with the MPCNC?

 

[jhitesma]

Are you going to cut carbon fiber with the MPCNC?

No immediately plans to...but eventually...I'll probably give it a try. Once I expand it out to 36"x48" for cutting DTFB sheets I'm not sure if it will still be rigid enough to do things like CF (which is why I'm kind of thinking about building a second already...I'd need a second set of motors, printed parts and hardware but could use the same electronics swapped between them to save some money.)

For now I'm very happy being able to cut acrylic, engrave and cut wood, and should be able to do things like mill simple PCB's. G10 shouldn't be an issue either, and I'm quite sure it would do CF - just slowly...but I'm not in a race and not doing things for production

Before I try CF though I'm looking to try some homemade composite materials for a rather unique frame idea I've hinted at a few times in the past

But making it bigger and cutting DTFB is my primary interest right now. I miss flying fixed wing and want to crank out a few more planes...but am too broke to buy speed build kits and too lazy to cut foam manually. I've had plans for the viggen, mini guinea, speedster, a replacement FT-22, replacement racer, and a few others that I printed and in some cases even glued to foam...but haven't got around to cutting out (mostly because every time I clear the counter to make room for it my wife clutters it again before I can get my cutting board out!)

Of course finishing the UBMQ is my #1 priority before I take the MPCNC apart to make it bigger

 

[jipp]

very cool. i want a 3d printer now.. you should get out a box of dollar tree foam. i know i have a box of the stuff here. anyhow and give it a go for a the small wing then you can 3d print a pod for it too ( the one im thinking of uses only one piece of foam i think.. i have the plans next to me to cut out one day.. ).. would be a fun project i think.. and good way to take advantage of the 3d printer too for the power pod..
last weekend i was in home depot getting some parts to replace my moms sink for her mothers day present. anyhow they had this square piece of pink foam about 3" thick by 2 and half feet square. anyhow first thing that came to my mind was body of a glider and for 5.00 not to expensive to try.. i still may try something like that as the summer days get longer and hotter, i could work on that in front of the a/c ..
anyhow could be fun on your machine that's for sue. good luck with carbon, i can only imagine that would be hell on tool bits.
chris.

 

[jhitesma]

You know what...having a CNC on your dining room table is pretty sweet

I ended up having plenty of time this evening to make a few chips with the new single flute plastic cutting 1/8" endmill that came in the mail today.

This bit cuts acrylic and polycarbonate (lexan) like butter. It's amazing. I also have a 1/16" version but I haven't had the guts to try it yet since it looks a lot more fragile. This one is a bit wasteful with it's large kerf...but plunging it gives a nearly perfect hole for 3mm bolts while the 1/16" one would have to carve out a hole. So there are pluses and minuses to either one.


Most of the mess there is just because I left the protective plastic on this sheet of lexan while cutting it to keep it clean. I stuck it to my spoil board with some carpet tape and it held great. The lexan measured 2.25mm so I went with a 2.5mm deep cut..but something must be a bit off on my Z because it just barely scratched the spoil board. Did make a clean up all the way through but I expected a bit more of a cut into the spoil board.



The engraving wasn't perfect but I didn't expect it to be. Engraving is best done with a pointed bit and while this bit does have a bit of an angle on it it's really not a good engraving bit. Too flat of a an angle and too big. But with the right font and keeping it at just 0.25mm deep it does make the top a little more interesting.

I also cut a new bottom plate, I couldn't think of anything really interesting to put on it...and it's going to be hidden under my battery anyway:



Bit of work with my hex driver and 5.5mm socket...and it's looking sharp!


I did find one small spot where the PETG center cracked just a bit. Not sure if it was from the crash I had on this setup that also cracked the 1/16" ply top or if it was from installing the arms because of the print being a bit tight. It's in a spot that doesn't give much strength so I'm not worried about it. As I keep saying most of the strength should come from the plates not the 3D angle section anyway.

So it's ready for some bashing! I'd like to flame polish the edges of the lexan to really finish it...but when the endmill cut through the carpet tape it smeared adhesive on the cut in some places and I don't have anything on hand that can get it off without risking damage to the lexan and I fear flame polishing with that on there would just make a mess. So instead I'll crash test it a bit tomorrow and this weekend before worrying about making it pretty If it survives this weekend then I'll take a few final steps to make it nice and take the design live on thingiverse.

 

[jipp]

ok, you need to do up a 130mm insect frame in lexcan. id go with a green tint.. man the options are boundless.. and at the 130mm size it would be strong even if it was just plexi id imagine.


chris.

 

[jhitesma]

Got to fly a pack on it tonight.

Good news. It survived a few decent hits into dirt and one smack into concrete with nothing but a few scuffs, a broken antenna, and a broken prop!

Bad news, probably won't have much video thanks to the neighbor kid trying to talk to me while I was flying again

 

[jipp]

sounds like the frame is gonna ork out ust fine for you> oohoo, have to love it hen a plan orks out> ( my keyboard is broke< im typing holding the shift key don or i get all caps< and the letter between q and e ont display and a couple of others it seems< guess i need to buy a ne keyboard sigh>> spilt pepsi on the keyboard sih. )

looks like video proof il ust be had tomoro. heh.

chris

 

[airhawk]

I cant be more excited for this frame if anything ive never been more excited im going to try to make an hk order to build up parts to start. Im going off of the hardware list you sent me to get the other stuff needed for the build.

 

[jhitesma]

No to pressure you or anything airhawk...but the sooner you get yours build (now that I finally sent it to you) the sooner I'll be inspired to make an FPV pod for it

In other news. Who's what in my GCS? Such autotune, much wow!



It's not exactly "hot news" as the DTFc has been out for a few weeks now at rotorgeeks but I've known about it for some time and dropped a few hints about it the past few months. It's shipping with betaflight...but I have no interest in actually flying betaflight. However DTF has been involved in dRonin since it started and the next release of dRonin (Scheduled for the start of July but will probably be a week or two late) should hopefully include full support of the DTFc. DTF put out a call for dRonin devs willing to help test, and while my only contributions so far have been a few lines of HTML adding the facebook and twitter pages to the GCS start screen - I have shown a willingness to test unproven code and lucked into helping isolate a few annoying bugs. So I was honored with a board to help test with

Right now there's no official build of dRonin available for the DTFc. The code is in a pull request on github but no binaries are available. So first thing I had to do was dust off my build environment and compile the FW/GCS. I got sidetracked with some windows updates and setting up a "new" computer (actually my ~8 year old vista machine from work that's been collecting dust for 3 years since it was replaced but which I dusted off and forced to accept windows 10 so I'd have a "modern" computer by my CNC machine in my old office which is transitioning into a workshop.) I actually got the build working a few days ago, but wasn't able to actually start GCS because this computer is old enough and it's video system is cheap enough that it won't run OpenGL2.1 since it isn't offiically supported under W10 Thankfully though icee pointed me to a work around (https://github.com/d-ronin/dRonin/wiki/Running-the-GCS-on-non-OpenGL-2.1-capable-systems) and I was able to get GCS up!

Then I had to do the initial flash...Since the DTFc comes with betaflight pre-loaded you can't just hook it up to GCS and flash it as there's no bootloader with betaflight. You have to use DFU mode to flash. I tried using the trick I came up with that's become standard practice for putting dRonin on Naze32 and use the CF configurator to flash it...but had no luck. I was able to get CF to recognize the board in DFU mode after using zadig to upgrade the drivers...and it could even connect...but it wouldn't flash my dRonin .hex And STM's DfUseDemo wouldn't detect the board after switching to libusb drivers with zadig...and I couldn't get the original STM drivers back. Well, actually this computer is just so slow I got frustrated and gave up trying to get them back. Instead today I used my other old computer which still has the STM drivers on it and was able to make a .dfu from the .hex and flash it with DfUseDemo.

Then went to connect it to GCS...and...couldn't get GCS to start on this ancient &!*$ computer again But this time it was my own fault. Too much time out in the sun this weekend leveling our yard for our daughters new pool apparently. I forgot that with custom builds on windows you have to start GCS from the command line unless you fully "install" it. Once icee reminded me of that - boom it all just worked and I got the doge welcoming me in the GCS


Now - before this tale of frustration and woe scares anyone away....remember I'm doing this kind of testing so end users won't have to! Plus most people aren't crazy enough to be running windows 10 on decade old hardware and compiling their own FC/GCS

So stay tuned. I'm too tired to fire up the soldering iron and put this thing into the UBMQ tonight...and even if I did I'm not sure which RX I'd use right now...I'm still kind of short on RX's. (But have 5 brotronics openLRS RX's waiting for me to solder them up and one I partly soldered waiting for me to fix a few remaining solder bridges due to too getting too much paste on it.) I have a few options but aren't super excited about any of them. Still I want to fly this thing so I'll have to make a choice soon.

Even so - having the FC and PDB integrated like this will help save a lot of space in the UMBQ and free up room for more RX options. So I am looking forward to getting it in there. And with the build in current/voltage monitoring I really want to either add a FPV pod with camera and micro minum HUD or use a RX option that will give me some telemetry. So stay tuned, this experimental build is getting a bit more experimental (or maybe just mental) soon!

 

[Jugsy]

Such a cool project, I'll be following this.

 

[jhitesma]

Glad to hear you like it Jugsy! I really need to get the files up on thingiverse. I've given it some good crashes and haven't done anything worse than break a prop...well...maybe damaged the USB on the FC today.

A week or two ago I got a mailer from surveilzone about a sale on DAL props...figured I may as well take advantage and finally give some tri blades a go. Got a couple sets each of the TJ5045's, T5045's, and T5045BN's to play with. Threw a set of the TJ5045's on today since they looked like the least aggressive and went out for a quick pack in the front yard.

Wasn't very happy. Definitely a loss of power compared to the 6045 FC's I've been running on this thing and didn't feel any more agile. And the tune was definitely non-optimal. Quad just felt loose and sloppy compared to what I'm used to. And the angle arms already feel a bit looser and sloppier than a flat quad to me so toss in my rusty thumbs from not flying much lately and it was a pretty sketchy flight.

Even so, wanted to give it a good thrash so did a few flips and rolls. Did ok, even doubles were tolerable. But then forgot how much less power the TJ5045's were giving me and smacked it into the ground hard enough to eject the battery. One blade on one prop was a bit bent and a good bit of sand in the motors...but overall no real damage done. Or so it seemed.

Came inside and realized I hadn't re-tuned for the different props. So went to hookup to GCS to configure for an autotune...and...realized I have't done that since upgrading this computer to windows 10. windows installed new drivers but the board keeps connecting and disconnecting constantly The USB cord is a VERY tight fit under the top plate (I had to shave down the connector housing to make it fit in fact) so I may have damaged the USB connector on the board. Will have to do a bit of investigating. I was still able to connect to the DTFc and a sparky1 with no issues so definitely seems to be this Sparky2's USB that's not happy

Well, I was planning on swapping the DTFc into here anyway!

As for the sparky2...assuming I can get the USB playing nice again I was planning on putting it into this new beast I started printing today:



Anyone care to guess what it is? It will be using a FC with dRonin...and that shroud is sized to fit a 5" prop...but I'm not sure what forum to post the build thread in because it's not fixed wing and not technically a multirotor

 

[jhitesma]

Did manage to catch a bit of video of it today without the neighborhood kids getting in the way:

Reviewing I remembered what I was doing when I crashed.

Bit of advice. Don't try testing hangtime LOS at low altitude...especially if it's been a few weeks since you last flew and just changed props without re-tuning

 

[cranialrectosis]

Your Sparky2 is not the problem. Windows 10 is not an upgrade.

I have been playing with the DAL 6045s and to me comparing to HQ 6045s it's like dropping a lipo cell. I break the DALs almost as often and the loss of power, particularly when recovering from a move, just hurts too much.

Put the new ducted fan drone in the Mad Men section.

You should consider an article if it flies.